Deep Learning-Based Biomimetic Identification Method for Mask Wearing Standardization

Abstract

Deep learning technology can automatically learn features from large amounts of data, with powerful feature extraction and pattern recognition capabilities, thereby improving the accuracy and efficiency of object detection. [The objective of this study]: In order to improve the accuracy and speed of mask wearing deep learning detection models in the post pandemic era, the [Problem this study aimed to resolve] was based on the fact that no research work has been reported on standardized detection models for mask wearing with detecting nose targets specially. [The topic and method of this study]: A mask wearing normalization detection model (towards the wearing style exposing the nose to outside, which is the most obvious characteristic of non-normalized style) based on improved YOLOv5s (You Only Look Once v5s is an object detection network model) was proposed. [The improved method of the proposed model]: The improvement design work of the detection model mainly includes (1) the BottleneckCSP (abbreviation of Bottleneck Cross Stage Partial) module was improved to a BottleneckCSP-MASK (abbreviation of Bottleneck Cross Stage Partial-MASK) module, which was utilized to replace the BottleneckCSP module in the backbone architecture of the original YOLOv5s model, which reduced the weight parameters' number of the YOLOv5s model while ensuring the feature extraction effect of the bonding fusion module. (2) An SE module was inserted into the proposed improved model, and the bonding fusion layer in the original YOLOv5s model was improved for better extraction of the features of mask and nose targets. [Results and validation]: The experimental results indicated that, towards different people and complex backgrounds, the proposed mask wearing normalization detection model can effectively detect whether people are wearing masks and whether they are wearing masks in a normalized manner. The overall detection accuracy was 99.3% and the average detection speed was 0.014 s/pic. Contrasted with original YOLOv5s, v5m, and v5l models, the detection results for two types of target objects on the test set indicated that the mAP of the improved model increased by 0.5%, 0.49%, and 0.52%, respectively, and the size of the proposed model compressed by 10% compared to original v5s model. The designed model can achieve precise identification for mask wearing behaviors of people, including not wearing a mask, normalized wearing, and wearing a mask non-normalized.

Keywords: BottleneckCSP; HSV space; artificial intelligence; mask; post-COVID-19 era; se module.

Figure 1

Situations of people wearing masks non-normalized.

Figure 2

Architecture of the original YOLOv5s network.

Figure 3

Focus module.

Figure 4

BottleneckCSP module.

Figure 5

Bottleneck module.

Figure 6

SPP module.

Figure 7

BottleneckCSP-MASK module (improved BottleneckCSP module).

Figure 8

Architecture of SE module.

Figure 9

Architecture of YOLOv5s-MASK network.

Figure 10

Overall flowchart of the proposed mask wearing normalization detection method.

Figure 11

Examples of people wearing masks normalized (a), wearing masks non-normalized (b), and not wearing masks (c) in datasets.

Figure 12

The original example image (a) and the resulting images after enhancement of 0.6 × V (b), 0.8 × V (c), 1.2 × V (d), and 1.6 × V (e).

Figure 13

Two examples (a,b) of generated training images utilizing Mosaic method.

Figure 14

Variation curves of precision, recall, and mAP indicators in model training.

Figure 15

Network training loss curve.

Figure 16

Recognition results for the three mask wearing styles not wearing mask, normalized wearing, and wearing mask non-normalized based on the YOLOv5s-MASK network.

Figure 16

Recognition results for the three mask wearing styles not wearing mask, normalized wearing, and wearing mask non-normalized based on the YOLOv5s-MASK network.

Figure 16

Recognition results for the three mask wearing styles not wearing mask, normalized wearing, and wearing mask non-normalized based on the YOLOv5s-MASK network.